Mapping Micro-Urban Heat Islands Using LANDSAT TM and a GIS

In this article, we look at a study that uses Geographic Information Systems (GIS) and LANDSAT TM (satellite data) to identify Micro-Urban Heat Islands (MUHIs) in Dallas, Texas, and to compare the spatial distribution of these MUHIs and tree density. Micro-Urban Heat Islands are "isolated urban locations that produce 'hot spots' within a city." Often these heat islands lead to a change in the climate of the area, and "most urban areas around the world with populations greater than 100,000 have heat islands that are 1-4.5 degrees (celsius) warmer than rural temperatures." One idea that has been suggested to reduce this effect of heat islands is to add more tree cover to urban areas.

The study mapped the MUHIs in the White Rock Lake area of Dallas using LANDSAT TM processing and classification, thermal map generation, and finally, a process of merging with GIS to produce the final map. Using LANDSAT TM, those conducting the study were able to create a thematic map which showed different pixel clusters, which were then grouped together to identify roads, bodies of water, buildings, grass, etc. Using the GIS layer that showed tree cover/density, the LANDSAT TM data was merged with tree cover to create the final map showing the distribution of tree cover compared to heat islands.

The figure above shows tree cover and MUHIs. This study concluded that heat islands are 5-11 degrees (celsius) warmer than surrounding areas by midmorning. The coolest areas within this area of Dallas were those with more tree cover, which suggests that increasing the density of trees within urban areas may be effective in decreasing the climate changes that come from MUHIs.

Aniello, C., Morgan, K., Busbey, A., & Newland, L. (1995). Mapping micro-urban heat islands using LANDSAT TM and a GIS. Computers & Geosciences,21(8), 965-969. doi:10.1016/0098-3004(95)00033-5